@Oohnorak

"Each path leading to a different next depth is a really bad idea."

I don't have much more than anecdote, but I do think that it could actually work pretty well.

To start, we've already got this in place where your options are to wait for the optimal level or to go straight ahead. In theory, this should cause a much higher degree of turmoil than the split path suggestion- do we visit the D24 compounds now? Or wait for the wolver den that lies just beyond it? This sort of scenario happens all the time. The same decision takes place whenever a danger room is encountered- do we use time and energy and crystal energy now for better resources, or do we keep going and ignore it? Finally, we already have split paths, but to a much smaller extent in levels like firestorm citadel and Ironclaw munitions factory.

The split path suggestion is always equal to or better than these scenarios, because it has less factors to consider and offers a more obvious choice. For instance, if each path is the same length and difficulty, it's a matter of preference. While there are a number of people that cannot handle such a monumental and inconsequential decision being made against their favor, I'd like to think that the bulk of knights are willing to compromise and have observed them doing so.

If the split path was a HUGE decision; i.e. each path provides different rewards, then the party would decide ahead of time and discuss where they were going and group up appropriately. Players also already do this whenever they log on and choose a run to join as well as during events like Apocrea and MOUNT DOOM: XMAS EDITION-- Apocrea can either act as a series of arena levels if you'd prefer radiant fire crystals, a source of tokens if you're lacking an obsidian piece that you want, or a source of whatever comes your way if you're all willing to compromise+need both while the evil Christmas event has multiple exits for varying levels of loot. It isn't uncommon for players on these levels to decide whether they're going all the way up the mountain/leaving at a specific point or in the case of Apocrea what strategy they'll be using on the arenas.

---------Strengths-----------

Lastly, the strengths of such a system are pretty enormous, should the developers choose to endorse them.

Imagine a set of three (T3 only) levels called "Gremlin Laboratory" that are always parallel to one another, each with a different themed level before it. These levels would be like functioning versions of the compounds we see normally.

-Gremlin Laboratory-

Level 1. Splitter level = features sign posts telling the knight what's ahead so they can choose, primarily gremlin guards and gremlins shipping crates around.

Levels 2+3 set A = Undead + gremlin; as if the gremlins were experimenting on the undead. Grave scarabs everywhere. 2 levels
Unique fight: 2 respawning gremlin menders, 3 waves of scarabs and gremlin knockers

Levels 2+3 set B = Slime + gremlin; gremlins doing science on slimes. Like a lesser version of C42 with glop drops everywhere. 2 levels
Unique fight: constant falling missiles with loads of glop drops + 1 mender on the last wave

Levels 2+3 set C = Beast + gremlin; gremlins breeding dust buns to use in combat against the Spiral Order. 2 levels
Unique fight: gremlin stalkers with respawning dust buns; lots of respawning explosives

Special drops from A/B/C = Unique gremlin schematic that becomes obsolete and disappears with the gate

Level 4. End level = A shorter fixed compound with extra gremlins instead of minis; randomly has 2-4 of 7 extra areas

Each area would contain a small number of tokens towards getting new dustbun/slime/beast themed battle sprites/reskinned battle sprites/gear, to make it all worthwhile should you play through the gate enough times. This kind of thing existing in the arcade and not the mission system is almost impossible right now, but could easily be done with branching paths. The purpose of it would be to drastically increase replayability and provide incentive to go on and master new and difficult areas, as well as to entertain.

In the example above, the first time the party went through they'd all have schematic A and at least 3 new schematics in a party of four would always be born, as the party could only open gate area A. The second time, the party would benefit most from going on the B path, as this would make for 3/7 areas the party could go on, and would end up with at most 4 A/4 B schematics and at least 1 A schematic and 2 B schematics remaining between them. The third time they went, the party would most likely go for area C, and would end up with at most 3 of 2 schematic and 4 of the last, and at least no remaining schematics (would require exactly A+C, B+C, A+B+C, and C, with 2 B and 2 A going in). But lets say I have 2C and 1A schematics going into a fresh party that hasn't been through yet. If we do B, then I can use my schematics to allow us to visit even more areas, but if we do A or C we gain less.

So the party gets to the splitter level or Haven or someplace and talks it over to decide an optimal path with optimal team members, and uses the schematics they've found to go down the best possible route. This sort of thing is communication and team work at its finest, and Spiral Knights is almost certainly a cooperative game.

The game can also just have a sign post whenever there's a branching path, or indicate it somehow. And go read it a third time, I fixed some stuff and added a second part. Maybe just read the second half of it.

Plus the current arcade setup has very little way to tell what level is coming up as is. Like, you really don't know until you get to the elevator whether it'll be an arena or what.

Yep. I only meant that there's no indication of how long the levels will be in place for and the decision only presents itself at the end of the level.

The larger party run has been an idea since forever, but it hasn't ever really been added. I wouldn't mind seeing a few select runs made as a sort of "boss rush" type mechanic, but it would have to be for some sort of an event. Most of the concerns fall back on the difficulty increase with more people in terms of monster health and the knight's ability to cope, which goes like so-

Number of knights...monster health
1....100% = base monster health
2....133% = 33% more
3....166% = Yes, this is linear
4....199%
5....232%
6....265%
7....298%
8....331%

Monster health per knight ratio-
1 knight = 1
2 knights = .66
3 knights = .55
4 knights = .49
...
8 knights = .41
Infiknight = .33

So the game would get much easier with a party of 8 in terms of how much damage is expected of each individual knight, but there is a sort of secondary effect that you would expect to see as well in that a party of 8 might happen to have a few weak links....

Knights performing at 100% of their effort = percent of monster health each knight is responsible for
8 = .41 = if everyone is effective, it works incredibly well
7 = .47
6 = .55 = if two knights are AFK/worthless, the difficulty is the same as having 3 knights
5 = .66 = Same efficiency as having 2...
4 = .82
3 = 1.1 = harder than solo
2 = 1.65
1 = 3.31 = You are now carrying seven people.

So as you can see, the math does come out to where the game would function with this number of people, but failure to communicate and inability to deal damage in such a fashion would be huge factors for a run like this-- especially since knights being hindered by other knights might reduce their effectiveness.. so if we had 2 AFK party members and were half as effective due to cutting each other off, all 6 active knights would be better off solo.

Using a strategy, the knights could go WAY above and beyond what is ever expected of them.. for instance, a party where each of the 8 members brought some combination of shivermist buster and the various autoguns would be just about unbeatable- in a party of 1, you might go shiver, bltiz, shiver, blitz.. but in a party of 3 + 1 shiver, you'd be going shiver/blitz/blitz/blitz, shiver/blitz/blitz/blutz.... which would mean that you'd be 1.5x more effective due to attacking 6x as often for 3 extra party members. The 8 shiver+blitz party described above would be able to go a step further and reach 14x the attacks a solo knight would have, making them 1.75 times as effective on top of their lower ratio of health.

The down side to this is that stepping on each other would be awful as well as commonplace. For instance, imagine if half the party tried to use brandishes and vortex bombs but a select few people brought weapons like blast bombs and pulsars to the fight. Only one group could really be fighting at a time, since the strategies cancel each other out, and dipping below an efficiency of 4 100% knights would be miserable.

Would it be worth giving a shot? Yeah sure, but only in a specific context that provides a basis for it. Like for the Christmas event next year- we could be given a huge arena with gremlins to fight in a party of 8. Maybe like the knights are ambushing the grinchlins or something. It would be cool and zany, but not serious enough to detriment the games via the problems outlined above.